Improving Fast and Frugal Modeling in Relation to Regression Analysis: Test of 3 Models for Medical Decision Making
2009 (English)In: Medical decision making, ISSN 0272-989X, E-ISSN 0112-0006, Vol. 29, no 1, 140-148 p.Article in journal (Refereed) Published
Background. A matching heuristic (MH) model of decision makinghas been evaluated previously in a series of studies on medicaldecision making. The authors' purpose is to evaluate an extendedMH model that considers the prevalence of cue values. Methods.Data from 2 previous studies were reanalyzed, one on judgmentsregarding drug treatment of hyperlipidemia and the other ondiagnosing heart failure. The original MH model and the extendedMH model were compared with logistic regression (LR) in termsof fit to actual judgments, number of cues, and the extent towhich the cues were consistent with clinical guidelines. Results. There was a slightly better fit with LR compared with MH. Theextended MH model gave a significantly better fit than the originalMH model in the drug treatment task. In the diagnostic task,the number of cues was significantly lower in the MH modelscompared to LR, whereas in the therapeutic task, LR could beless or more frugal than the matching heuristic models dependingon the significance level chosen for inclusion of cues. Forthe original MH model, but not for the extended MH model orLR, the most important cues in the drug treatment task wereoften used in a direction contrary to treatment guidelines.Conclusions. The extended MH model represents an improvementin that prevalence of cue values is adequately taken into account,which in turn may result in better fit and in better agreementwith medical guidelines in the evaluation of cues.
Place, publisher, year, edition, pages
Sage Publications , 2009. Vol. 29, no 1, 140-148 p.
medical decision making, fast and frugal, matching heuristic, judgment analysis
Research subject Psychology
IdentifiersURN: urn:nbn:se:su:diva-35325DOI: 10.1177/0272989X08326091OAI: oai:DiVA.org:su-35325DiVA: diva2:286991
Financial support for this study was provided entirely by a grant from Stockholm County Council.2010-01-152010-01-152015-09-10Bibliographically approved